Overview In this role you lead and shape the science strategy for agentic AI in Excel, translating frontier AI research into reliable product capabilities. You manage a multidisciplinary team and own evaluation, release-readiness, and improvements across quality, latency, reliability, safety, and cost. You partner with engineering, product, design, and AI organizations to turn research into customer value. You'll drive cross-functional collaboration and build a resilient, high-impact team dedicated to advancing Excel's AI experiences.
Responsibilities- Lead, grow, and mentor a multidisciplinary team of researchers and applied scientists
- Set the research and applied-science strategy for agentic AI in Excel
- Drive scientific direction for agent architecture, model and tool selection, prompts, skills, planning, and human-agent collaboration
- Define how agent quality is measured through benchmarks, graders, regression suites, and multi-turn evaluations
- Establish release-readiness criteria for new models and agent capabilities
- Lead evaluation and integration of frontier foundation models balancing quality, latency, reliability, safety, capacity, and cost
- Diagnose agent failures using trajectory analysis and customer-representative scenarios
- Translate research prototypes into production-ready capabilities in partnership with engineering, product management, and design
- Build operating mechanisms for model comparisons, quality reviews, and technical decision-making
- Develop partnerships with internal model teams, research organizations, and external AI labs
- Communicate findings to senior leadership and support external research engagement through publications and conferences
- Recruit, onboard, and develop research talent while establishing durable technical leadership
Key requirements- Bachelor's/Master's/Doctorate in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field with required years of experience
- 3+ years of people management experience
- leadership and people management
- executive communication
- cross-functional collaboration
- LLMs and AI agents
- agent architectures and prompting
- evaluation frameworks and benchmarks
